2. Measurements . 



Field experimentation was emphasized with support as needed from 

 the laboratory, phytotron, and greenhouse. Most variables were 

 tested in replicated trials; exploratory and field-scale plantings 

 were not replicated. 



The following measures were used to estimate plant growth and 

 development in the field: 



(a) Aerial dry weight. Oven dry weight of abovegrouni growth 

 per transplant (hill) or per unit area. 



(b) Belowground dry weight (rhizomes and roots). Oven dry 

 weight of rhizomes and roots of first-year transplants; for older 

 plantings two cores were usually taken from each quadrat harvested 

 for aboveground growth. Samples were taken 30 centimeters deep 

 with a stainless steel coring tube 8.5 centimeters in diameter. 

 Samples were washed and plant parts were separated, weighed, and 

 recorded as belowground growth per hill or per unit area. 



(c) Number of flowers. Number of flowering culms per hill 

 or per unit area. 



(d) Number of center culms. Number of culms (stems) cluster- 

 ed around the original transplant (recorded for first-year trans- 

 plants only) . 



(e) Number of rhizome culms. Number of culms growing from 

 rhizomes away from the original transplant (spread: recorded on 

 first-year transplants only) . 



(f) Height. Distance from base to tip of culm in centimeters 

 (usually the average of five culms per transplant) . 



(g) Basal area. Area covered by culms at ground level. This 

 was determined by measuring the cross-sectional diameter of tight- 

 ly bunched culms; data were recorded per hill or per unit area. 



The exposed location and unstable conditions of many shoreline 

 sites make them more vulnerable to damage or total loss than is typi- 

 cal of upland studies; higher experimental errors are usually obtained 

 from coastal studies. Compensation for this problem was attempted 

 through duplication of field tests, wherever possible. Hazards to new 

 plants were: Erosion or sand deposition due to wave action, under- 

 mining resulting from channel shifting, and burial by windblown sand. 



Growth data were collected near the end of each growing season. 

 Some nondestructive measurements were taken during the first week in 



13 



